Disaggregating state capacity: explaining policy effectiveness in Latin America, 1996-2006

In this thesis, I use the concept of state capacity to account for cross-national levels of variation in policy effectiveness in Latin America between 1996 and 2006. In doing so, I make four contributions to the literature. Firstly, I develop a theoretically-grounded conceptualization and an empirically systematic measurement of policy effectiveness for 18 Latin American countries along security, welfare and economic policy dimensions. Secondly, I develop a novel conceptualization and operationalization of state capacity along three key dimensions – infrastructural power, bureaucratic capacity and political capacity. By disaggregating state capacity into these three distinct (but mutually reinforcing) constituent dimensions, I integrate different strands of the literature on state capacity and purport to increase the explanatory power of state capacity as a conceptual variable. Thirdly, I develop a simple but theoretically differentiated model of policy effectiveness which maps out and incorporates different kinds of politics-centred explanations of effectiveness, while situating state capacity as a direct (but not exclusive) cause of effectiveness. The final contribution is empirical: I triangulate statistical methods, crisp-set qualitative comparative analysis and an extended qualitative comparison of two countries (Argentina and Chile) along three policy areas (health, citizen security and economic regulation) to provide a rich analysis of the influence the different dimensions of state capacity have on each policy effectiveness dimension. My results suggest, firstly, that state capacity differences are indeed large and important for explaining within-region differences in effectiveness; secondly, that the use of this disaggregated approach provides important theoretical and empirical payoffs for understanding the multiple ways in which states affect outcomes; and thirdly, that differentiating ‘institutions as organizations’ (such as the state) from the standard understanding of ‘institutions as rules’ allows us to improve on standard institutionalist accounts of the influence of politics on development.